Go to a local shop and have them bend some 1/8" aluminium in a U shape. Some timber on the ends and a couple of aluminum angle will finish it. Easy, cheap and locally made. Have a look at the pictures of my 6BQ5 that I build nearly 7 years ago and that I posted today in the transformer thread.
Wow, thanks for the replies everyone!
Making my own PCB is definitely an option for me, I used to use packages like Altium and EAGLE for work, but EasyEDA is shaping up to be really nice, and runs in the browser.
PSU design may be a challenge. I've designed low voltage 'buck' type converters before, so I'm eyeing up an SMPS, but I also don't want to get bogged down with side projects. Benchtop PSUs exist for a reason.
I've downloaded the RCA radiotron manual from the sticky.
As a first attempt, I'll probably go easy on myself with a hybrid tube gain stage and solid-state power stage.
I also have some ideas for future projects with electro-magnet based speakers. With access to 2 air-gapped coils, there could be some interesting possibilities to linearise them, unlike with the standard system of permanent magnets. Agonising over perfecting the magnetic gap seems misguided when what really seems to be missing is precise feedback. Probably off-topic but it might be interesting to put it all together into an active speaker.
Making my own PCB is definitely an option for me, I used to use packages like Altium and EAGLE for work, but EasyEDA is shaping up to be really nice, and runs in the browser.
PSU design may be a challenge. I've designed low voltage 'buck' type converters before, so I'm eyeing up an SMPS, but I also don't want to get bogged down with side projects. Benchtop PSUs exist for a reason.
I've downloaded the RCA radiotron manual from the sticky.
As a first attempt, I'll probably go easy on myself with a hybrid tube gain stage and solid-state power stage.
I also have some ideas for future projects with electro-magnet based speakers. With access to 2 air-gapped coils, there could be some interesting possibilities to linearise them, unlike with the standard system of permanent magnets. Agonising over perfecting the magnetic gap seems misguided when what really seems to be missing is precise feedback. Probably off-topic but it might be interesting to put it all together into an active speaker.
I'd go with a Tubelab SSE. George's designs are well thought-out and allows for plenty of tweaking should you so desire.
Alternatively, I'd design something around a sweep tube such as 6LR8/6LU8.
If you want to design your own, I'd start by getting a copy of Morgan Jones, "Valve Amplifiers".
Tom
Alternatively, I'd design something around a sweep tube such as 6LR8/6LU8.
If you want to design your own, I'd start by getting a copy of Morgan Jones, "Valve Amplifiers".
Tom
Here are some experiences with valve amplifiers. My first amplifier was a Mullard 3-3 with an EL84, followed by a pair of Mullard 5-10's with EL84s, Mullard 5-20's with EL34s. The reason was simple, detailed information was published in a Mullard book. This goes back 60 years; I was 14. These amplifiers were built over about the next 7 years. Later the 5-20's became Williamsons, eventually with KT88's and bigger output transformers. You will over time try different designs. The reason for changing to Williamson amps was the sound with class A triodes was much more enjoyable.
You may use different tubes to EF86, 6SN7 etc. but Class A triodes sound the most musical.
You may use different tubes to EF86, 6SN7 etc. but Class A triodes sound the most musical.
I'd start with a tube that's designed for audio. Sweep tubes can perform excellent, however, it is much harder to find information about them. For someone who is newer to tubes, it is much easier to design around something like a 6L6 or EL34. Plate curves are available and there are lots of existing designs to reference. Sweep tubes also don't take UL service well due to their low screen voltage ratings.
Also, I really recommend using new-production tubes for a first build. You can buy matched pairs and quartets that are already burned in for 24 hours. This doesn't guarantee that you won't have tube problems, but it certainly helps.
As another note, class A triodes may sound nice (I won't take a stance on this), but they also make less power than any other solution. Keep this in mind as you're choosing a design for your application. If you have 87 dB speakers, a medium size room and a love for rock music, you will probably be less than satisfied with a class A 2A3 amplifier.
Spot on! Paul Joppa has provided us with a useful rule of thumb for mating amps and speakers. The rule states that in "typical" listening spaces an amp speaker combo should be capable of producing 102 dB. SPL levels at a 1 M. distance. Applying Joppa's Rule to that 87 dB. situation yields a 32 WPC requirement. Both 3.5 W. 2A3s and 8 W. 300Bs rate to be highly unsatisfactory.
Abstract, according to your post #22 you are planning an active speaker with solid state power stage and high voltage vacuum tube gain stage. I've built exactly this type of speaker 15 years ago. The power transformers are inside the speaker at the bottom, the tube is a plain 6N3 triode gain stage (both sections are connected in parallel and powered by a high voltage power supply), the power stage is a chip amp bolted to the underside of the machined aluminium heatsink, the speaker port is on the back of the cabinet. The cabinet itself was a China sourced kit (Edfier maybe). I regretted this choice almost immediately, for several reasons. The main one: I quickly realised that strapping a triode to a solid state amplifier did not meet my goal. I like the sound of a full vacuum tube amplifier more. Then the cabinet was not the right shape for the driver I had, and due to the extra complexity of the built-in amplifier, the modification would not have been cheap or easy. Finally, the toroidal transformer had a little bit of hum but the placement inside the speaker made it much more noticeable. If you want to try a hybrid amplifier because you feel that a full vacuum tube amplifier is more difficult to build, according to my experience this is not true. On a hybrid amplifier, if you use the correct tube high voltages, you need two power supplies and you must use extra care to avoid damage to the semiconductor part. You also need a PCB for the solid state power stage. Compare that to the part count of many classic tube amplifiers: it is so low that you may even avoid the PCB completely. As example, there is a low power ECL86 / PCL86 single ended amplifier design with the same complexity level as many chip amps, that has been built countless times since the '60 when Mullard designed it, and still gives a good taste of the sound of a full tube amplifier.
Attachments
https://www.diyaudio.com/forums/tub...t-xfmrs-current-suppliers-13.html#post6088752
The small arctic white 6BQ5 amplifier
I'm targeting +/-5V swing in my MOSFET build ( Single Ended Class A MOSFET Current Amplifier - lechaudio ), and that's based on looking at the output voltage (using a scope) for an existing amplifier with music running at various levels. The system has 2-way 10" + 3.5" speakers, bi-amped with 10-20W class-D modules x 4ch, and parametric EQ with a PC sound card as the source.
A normal background level for my needs had peaks around +/-0.5-1V, and with louder tests I was able to estimate that a +/-5V soft limit should be more than adequate for me. It's a small living space.
OTOH, in simulations, THD readings were starting to fall below 0.05% @1kHz, 10mV in (250-300mV out), gradually rising to ~1% near clipping (200-300mV in, ~6V out). What I don't know yet is how that will translate to actual listening -- whether the limit will be clipping, or somewhere below that. I'll be tuning by ear mostly, so I won't have hard numbers, but when I hear what's happening, then I'll have a better feel for what needs to be changed.
I'm trying to figure out what the "tube sound" really means. My pet theory at the moment is that a high damping factor is the main culprit. Once real-world delayed energy from the speaker and box (or horn, or dipole baffle, etc) are factored in, 'negative' feedback must become positive feedback at some resonant modes where the reflections are 180 deg out of phase.
An awful lot of any (tube or SS) amp's sound is design dependent. The H/K Cit. 2 has a pretty high damping factor and neutral voicing. The "Deuce" beats many a SS amp at their own game of bass performance.
JMO, well executed tube circuitry introduces a tiny amount of 2nd order harmonic distortion that the human ear finds euphonic, while avoiding odd order products that ear "detests".
The vacuum triode is the most linear voltage amplifying device developed, to date. The specific triode found in 6SN7s and other types meets HIFI distortion criteria, without benefit of any sort of NFB.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.